rethink your priors
What is MCP, and is it complicated?
The short answer
MCP, the Model Context Protocol, is an open standard, introduced by Anthropic in 2024, for connecting AI applications to tools and data. Think of it as a power strip: every tool gets the same plug, so you don't wire each one by hand. The plug is the easy part. The real decision is what the AI is allowed to touch. Reading your calendar is one thing. Sending email, moving money or deleting records is another. Before connecting anything, list which actions are safe to run without asking, which need a person to approve, and which should never be available at all. Then connect only what the outcome needs. Most of the work in an MCP setup isn't plumbing. It's deciding, in advance, what you'd be comfortable with the AI doing while nobody is watching. Start with read-only access and add permissions one at a time.
Outcome first
- Decide what the AI may touch.
- The plug is the easy part.
Want this applied to your actual workflow? That part isn't a cartoon. Work with Eduardo. Longer writing: The Compounding Founder.